aboutsummaryrefslogtreecommitdiff
path: root/locale/programs/ld-collate.c
blob: 42fd6010645c733bb7ed9ca52ab732843a6b992c (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
/* Copyright (C) 1995, 1996, 1997, 1998, 1999 Free Software Foundation, Inc.
   This file is part of the GNU C Library.
   Contributed by Ulrich Drepper <drepper@gnu.org>, 1995.

   The GNU C Library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Library General Public License as
   published by the Free Software Foundation; either version 2 of the
   License, or (at your option) any later version.

   The GNU C Library is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   Library General Public License for more details.

   You should have received a copy of the GNU Library General Public
   License along with the GNU C Library; see the file COPYING.LIB.  If not,
   write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
   Boston, MA 02111-1307, USA.  */

#ifdef HAVE_CONFIG_H
# include <config.h>
#endif

#include <errno.h>
#include <error.h>
#include <stdlib.h>
#include <wchar.h>

#include "charmap.h"
#include "localeinfo.h"
#include "linereader.h"
#include "locfile.h"
#include "localedef.h"

/* Uncomment the following line in the production version.  */
/* #define NDEBUG 1 */
#include <assert.h>

#define obstack_chunk_alloc malloc
#define obstack_chunk_free free

/* Forward declaration.  */
struct element_t;

/* Data type for list of strings.  */
struct section_list
{
  struct section_list *next;
  /* Name of the section.  */
  const char *name;
  /* First element of this section.  */
  struct element_t *first;
  /* Last element of this section.  */
  struct element_t *last;
  /* These are the rules for this section.  */
  enum coll_sort_rule *rules;
};

struct element_t;

struct element_list_t
{
  /* Number of elements.  */
  int cnt;

  struct element_t **w;
};

/* Data type for collating element.  */
struct element_t
{
  const char *name;

  const char *mbs;
  const uint32_t *wcs;
  int order;

  struct element_list_t *weights;

  /* Where does the definition come from.  */
  const char *file;
  size_t line;

  /* Which section does this belong to.  */
  struct section_list *section;

  /* Predecessor and successor in the order list.  */
  struct element_t *last;
  struct element_t *next;
};

/* Special element value.  */
#define ELEMENT_ELLIPSIS2	((struct element_t *) 1)
#define ELEMENT_ELLIPSIS3	((struct element_t *) 2)
#define ELEMENT_ELLIPSIS4	((struct element_t *) 3)

/* Data type for collating symbol.  */
struct symbol_t
{
  /* Point to place in the order list.  */
  struct element_t *order;

  /* Where does the definition come from.  */
  const char *file;
  size_t line;
};


/* The real definition of the struct for the LC_COLLATE locale.  */
struct locale_collate_t
{
  int col_weight_max;
  int cur_weight_max;

  /* List of known scripts.  */
  struct section_list *sections;
  /* Current section using definition.  */
  struct section_list *current_section;
  /* There always can be an unnamed section.  */
  struct section_list unnamed_section;
  /* To make handling of errors easier we have another section.  */
  struct section_list error_section;

  /* Number of sorting rules given in order_start line.  */
  uint32_t nrules;

  /* Start of the order list.  */
  struct element_t *start;

  /* The undefined element.  */
  struct element_t undefined;

  /* This is the cursor for `reorder_after' insertions.  */
  struct element_t *cursor;

  /* This value is used when handling ellipsis.  */
  struct element_t ellipsis_weight;

  /* Known collating elements.  */
  hash_table elem_table;

  /* Known collating symbols.  */
  hash_table sym_table;

  /* Known collation sequences.  */
  hash_table seq_table;

  struct obstack mempool;

  /* The LC_COLLATE category is a bit special as it is sometimes possible
     that the definitions from more than one input file contains information.
     Therefore we keep all relevant input in a list.  */
  struct locale_collate_t *next;
};


/* We have a few global variables which are used for reading all
   LC_COLLATE category descriptions in all files.  */
static int nrules;


static struct section_list *
make_seclist_elem (struct locale_collate_t *collate, const char *string,
		   struct section_list *next)
{
  struct section_list *newp;

  newp = (struct section_list *) obstack_alloc (&collate->mempool,
						sizeof (*newp));
  newp->next = next;
  newp->name = string;
  newp->first = NULL;

  return newp;
}


static struct element_t *
new_element (struct locale_collate_t *collate, const char *mbs,
	     const uint32_t *wcs, const char *name, size_t namelen)
{
  struct element_t *newp;

  newp = (struct element_t *) obstack_alloc (&collate->mempool,
					     sizeof (*newp));
  newp->name = name == NULL ? NULL : obstack_copy (&collate->mempool,
						   name, namelen);
  newp->mbs = mbs;
  if (wcs != NULL)
    {
      size_t nwcs = wcslen ((wchar_t *) wcs) + 1;
      uint32_t zero = 0;
      obstack_grow (&collate->mempool, wcs, nwcs * sizeof (uint32_t));
      obstack_grow (&collate->mempool, &zero, sizeof (uint32_t));
      newp->wcs = (uint32_t *) obstack_finish (&collate->mempool);
    }
  else
    newp->wcs = NULL;
  newp->order = 0;

  /* Will be allocated later.  */
  newp->weights = NULL;

  newp->file = NULL;
  newp->line = 0;

  newp->section = NULL;

  newp->last = NULL;
  newp->next = NULL;

  return newp;
}


static struct symbol_t *
new_symbol (struct locale_collate_t *collate)
{
  struct symbol_t *newp;

  newp = (struct symbol_t *) obstack_alloc (&collate->mempool, sizeof (*newp));

  newp->order = NULL;

  newp->file = NULL;
  newp->line = 0;

  return newp;
}


/* Test whether this name is already defined somewhere.  */
static int
check_duplicate (struct linereader *ldfile, struct locale_collate_t *collate,
		 struct charmap_t *charmap, struct repertoire_t *repertoire,
		 const char *symbol, size_t symbol_len)
{
  void *ignore = NULL;

  if (find_entry (&charmap->char_table, symbol, symbol_len, &ignore) == 0)
    {
      lr_error (ldfile, _("`%s' already defined in charmap"), symbol);
      return 1;
    }

  if (find_entry (&repertoire->char_table, symbol, symbol_len, &ignore) == 0)
    {
      lr_error (ldfile, _("`%s' already defined in repertoire"), symbol);
      return 1;
    }

  if (find_entry (&collate->sym_table, symbol, symbol_len, &ignore) == 0)
    {
      lr_error (ldfile, _("`%s' already defined as collating symbol"), symbol);
      return 1;
    }

  if (find_entry (&collate->elem_table, symbol, symbol_len, &ignore) == 0)
    {
      lr_error (ldfile, _("`%s' already defined as collating element"),
		symbol);
      return 1;
    }

  return 0;
}


/* Read the direction specification.  */
static void
read_directions (struct linereader *ldfile, struct token *arg,
		 struct charmap_t *charmap, struct repertoire_t *repertoire,
		 struct locale_collate_t *collate)
{
  int cnt = 0;
  int max = nrules ?: 10;
  enum coll_sort_rule *rules = calloc (max, sizeof (*rules));
  int warned = 0;

  while (1)
    {
      int valid = 0;

      if (arg->tok == tok_forward)
	{
	  if (rules[cnt] & sort_backward)
	    {
	      if (! warned)
		{
		  lr_error (ldfile, _("\
%s: `forward' and `backward' are mutually excluding each other"),
			    "LC_COLLATE");
		  warned = 1;
		}
	    }
	  else if (rules[cnt] & sort_forward)
	    {
	      if (! warned)
		{
		  lr_error (ldfile, _("\
%s: `%s' mentioned twice in definition of weight %d"),
			    "LC_COLLATE", "forward", cnt + 1);
		}
	    }
	  else
	    rules[cnt] |= sort_forward;

	  valid = 1;
	}
      else if (arg->tok == tok_backward)
	{
	  if (rules[cnt] & sort_forward)
	    {
	      if (! warned)
		{
		  lr_error (ldfile, _("\
%s: `forward' and `backward' are mutually excluding each other"),
			    "LC_COLLATE");
		  warned = 1;
		}
	    }
	  else if (rules[cnt] & sort_backward)
	    {
	      if (! warned)
		{
		  lr_error (ldfile, _("\
%s: `%s' mentioned twice in definition of weight %d"),
			    "LC_COLLATE", "backward", cnt + 1);
		}
	    }
	  else
	    rules[cnt] |= sort_backward;

	  valid = 1;
	}
      else if (arg->tok == tok_position)
	{
	  if (rules[cnt] & sort_position)
	    {
	      if (! warned)
		{
		  lr_error (ldfile, _("\
%s: `%s' mentioned twice in definition of weight %d in category `%s'"),
			    "LC_COLLATE", "position", cnt + 1);
		}
	    }
	  else
	    rules[cnt] |= sort_position;

	  valid = 1;
	}

      if (valid)
	arg = lr_token (ldfile, charmap, repertoire);

      if (arg->tok == tok_eof || arg->tok == tok_eol || arg->tok == tok_comma
	  || arg->tok == tok_semicolon)
	{
	  if (! valid && ! warned)
	    {
	      lr_error (ldfile, _("%s: syntax error"), "LC_COLLATE");
	      warned = 1;
	    }

	  /* See whether we have to increment the counter.  */
	  if (arg->tok != tok_comma && rules[cnt] != 0)
	    ++cnt;

	  if (arg->tok == tok_eof || arg->tok == tok_eol)
	    /* End of line or file, so we exit the loop.  */
	    break;

	  if (nrules == 0)
	    {
	      /* See whether we have enough room in the array.  */
	      if (cnt == max)
		{
		  max += 10;
		  rules = (enum coll_sort_rule *) xrealloc (rules,
							    max
							    * sizeof (*rules));
		  memset (&rules[cnt], '\0', (max - cnt) * sizeof (*rules));
		}
	    }
	  else
	    {
	      if (cnt == nrules)
		{
		  /* There must not be any more rule.  */
		  if (! warned)
		    {
		      lr_error (ldfile, _("\
%s: too many rules; first entry only had %d"),
				"LC_COLLATE", nrules);
		      warned = 1;
		    }

		  lr_ignore_rest (ldfile, 0);
		  break;
		}
	    }
	}
      else
	{
	  if (! warned)
	    {
	      lr_error (ldfile, _("%s: syntax error"), "LC_COLLATE");
	      warned = 1;
	    }
	}

      arg = lr_token (ldfile, charmap, repertoire);
    }

  if (nrules == 0)
    {
      /* Now we know how many rules we have.  */
      nrules = cnt;
      rules = (enum coll_sort_rule *) xrealloc (rules,
						nrules * sizeof (*rules));
    }
  else
    {
      if (cnt < nrules)
	{
	  /* Not enough rules in this specification.  */
	  if (! warned)
	    lr_error (ldfile, _("%s: not enough sorting rules"), "LC_COLLATE");

	  do
	    rules[cnt] = sort_forward;
	  while (++cnt < nrules);
	}
    }

  collate->current_section->rules = rules;
}


static struct element_t *
find_element (struct linereader *ldfile, struct locale_collate_t *collate,
	      const char *str, size_t len, uint32_t *wcstr)
{
  struct element_t *result = NULL;

  /* Search for the entries among the collation sequences already define.  */
  if (find_entry (&collate->seq_table, str, len, (void **) &result) != 0)
    {
      /* Nope, not define yet.  So we see whether it is a
         collation symbol.  */
      void *ptr;

      if (find_entry (&collate->sym_table, str, len, &ptr) == 0)
	{
	  /* It's a collation symbol.  */
	  struct symbol_t *sym = (struct symbol_t *) ptr;
	  result = sym->order;

	  if (result == NULL)
	    result = sym->order = new_element (collate, NULL, NULL, NULL, 0);
	}
      else if (find_entry (&collate->elem_table, str, len,
			   (void **) &result) != 0)
	{
	  /* It's also no collation element.  So it is an character
	     element defined later.  */
	  result = new_element (collate, NULL, NULL, str, len);
	  if (result != NULL)
	    /* Insert it into the sequence table.  */
	    insert_entry (&collate->seq_table, str, len, result);
	}
    }

  return result;
}


static void
unlink_element (struct locale_collate_t *collate)
{
  if (collate->cursor->next != NULL)
    collate->cursor->next->last = collate->cursor->last;
  if (collate->cursor->last != NULL)
    collate->cursor->last->next = collate->cursor->next;
  collate->cursor = collate->cursor->last;
}


static void
insert_weights (struct linereader *ldfile, struct element_t *elem,
		struct charmap_t *charmap, struct repertoire_t *repertoire,
		struct locale_collate_t *collate, enum token_t ellipsis)
{
  int weight_cnt;
  struct token *arg;

  /* Initialize all the fields.  */
  elem->file = ldfile->fname;
  elem->line = ldfile->lineno;
  elem->last = collate->cursor;
  elem->next = collate->cursor ? collate->cursor->next : NULL;
  elem->weights = (struct element_list_t *)
    obstack_alloc (&collate->mempool, nrules * sizeof (struct element_list_t));
  memset (elem->weights, '\0', nrules * sizeof (struct element_list_t));

  if (collate->current_section->first == NULL)
    collate->current_section->first = elem;
  if (collate->current_section->last == collate->cursor)
    collate->current_section->last = elem;

  collate->cursor = elem;

  weight_cnt = 0;

  arg = lr_token (ldfile, charmap, repertoire);
  do
    {
      if (arg->tok == tok_eof || arg->tok == tok_eol)
	break;

      if (arg->tok == tok_ignore)
	{
	  /* The weight for this level has to be ignored.  We use the
	     null pointer to indicate this.  */
	  elem->weights[weight_cnt].w = (struct element_t **)
	    obstack_alloc (&collate->mempool, sizeof (struct element_t *));
	  elem->weights[weight_cnt].w[0] = NULL;
	  elem->weights[weight_cnt].cnt = 1;
	}
      else if (arg->tok == tok_bsymbol)
	{
	  struct element_t *val = find_element (ldfile, collate,
						arg->val.str.startmb,
						arg->val.str.lenmb,
						arg->val.str.startwc);

	  if (val == NULL)
	    break;

	  elem->weights[weight_cnt].w = (struct element_t **)
	    obstack_alloc (&collate->mempool, sizeof (struct element_t *));
	  elem->weights[weight_cnt].w[0] = val;
	  elem->weights[weight_cnt].cnt = 1;
	}
      else if (arg->tok == tok_string)
	{
	  /* Split the string up in the individual characters and put
	     the element definitions in the list.  */
	  const char *cp = arg->val.str.startmb;
	  int cnt = 0;
	  struct element_t *charelem;
	  void *base = obstack_base (&collate->mempool);

	  if (*cp == '\0')
	    {
	      lr_error (ldfile, _("%s: empty weight string not allowed"),
			"LC_COLLATE");
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  do
	    {
	      if (*cp == '<')
		{
		  /* Ahh, it's a bsymbol.  That's what we want.  */
		  const char *startp = cp;

		  while (*++cp != '>')
		    {
		      if (*cp == ldfile->escape_char)
			++cp;
		      if (*cp == '\0')
			{
			  /* It's a syntax error.  */
			  obstack_free (&collate->mempool, base);
			  goto syntax;
			}
		    }

		    charelem = find_element (ldfile, collate, startp,
					     cp - startp, NULL);
		    ++cp;
		}
	      else
		{
		  /* People really shouldn't use characters directly in
		     the string.  Especially since it's not really clear
		     what this means.  We interpret all characters in the
		     string as if that would be bsymbols.  Otherwise we
		     would have to match back to bsymbols somehow and this
		     is also not what people normally expect.  */
		  charelem = find_element (ldfile, collate, cp++, 1, NULL);
		}

	      if (charelem == NULL)
		{
		  /* We ignore the rest of the line.  */
		  lr_ignore_rest (ldfile, 0);
		  break;
		}

	      /* Add the pointer.  */
	      obstack_ptr_grow (&collate->mempool, charelem);
	      ++cnt;
	    }
	  while (*cp != '\0');

	  /* Now store the information.  */
	  elem->weights[weight_cnt].w = (struct element_t **)
	    obstack_finish (&collate->mempool);
	  elem->weights[weight_cnt].cnt = cnt;

	  /* We don't need the string anymore.  */
	  free (arg->val.str.startmb);
	}
      else if (ellipsis != tok_none
	       && (arg->tok == tok_ellipsis2
		   || arg->tok == tok_ellipsis3
		   || arg->tok == tok_ellipsis4))
	{
	  /* It must be the same ellipsis as used in the initial column.  */
	  if (arg->tok != ellipsis)
	    lr_error (ldfile, _("\
%s: weights must use the same ellipsis symbol as the name"),
		      "LC_COLLATE");

	  /* The weight for this level has to be ignored.  We use the
	     null pointer to indicate this.  */
	  elem->weights[weight_cnt].w = (struct element_t **)
	    obstack_alloc (&collate->mempool, sizeof (struct element_t *));
	  elem->weights[weight_cnt].w[0] = ELEMENT_ELLIPSIS2;
	  elem->weights[weight_cnt].cnt = 1;
	}
      else
	{
	syntax:
	  /* It's a syntax error.  */
	  lr_error (ldfile, _("%s: syntax error"), "LC_COLLATE");
	  lr_ignore_rest (ldfile, 0);
	  break;
	}

      arg = lr_token (ldfile, charmap, repertoire);
      /* This better should be the end of the line or a semicolon.  */
      if (arg->tok == tok_semicolon)
	/* OK, ignore this and read the next token.  */
	arg = lr_token (ldfile, charmap, repertoire);
      else if (arg->tok != tok_eof && arg->tok != tok_eol)
	{
	  /* It's a syntax error.  */
	  lr_error (ldfile, _("%s: syntax error"), "LC_COLLATE");
	  lr_ignore_rest (ldfile, 0);
	  break;
	}
    }
  while (++weight_cnt < nrules);

  if (weight_cnt < nrules)
    {
      /* This means the rest of the line uses the current element as
	 the weight.  */
      do
	{
	  elem->weights[weight_cnt].w = (struct element_t **)
	    obstack_alloc (&collate->mempool, sizeof (struct element_t *));
	  elem->weights[weight_cnt].w[0] = elem;
	  elem->weights[weight_cnt].cnt = 1;
	}
      while (++weight_cnt < nrules);
    }
  else
    {
      if (arg->tok == tok_ignore || arg->tok == tok_bsymbol)
	{
	  /* Too many rule values.  */
	  lr_error (ldfile, _("%s: too many values"), "LC_COLLATE");
	  lr_ignore_rest (ldfile, 0);
	}
      else
	lr_ignore_rest (ldfile, arg->tok != tok_eol && arg->tok != tok_eof);
    }
}


static void
insert_value (struct linereader *ldfile, struct token *arg,
	      struct charmap_t *charmap, struct repertoire_t *repertoire,
	      struct locale_collate_t *collate)
{
  /* First find out what kind of symbol this is.  */
  struct charseq *seq;
  uint32_t wc;
  struct element_t *elem = NULL;

  /* Try to find the character in the charmap.  */
  seq = charmap_find_value (charmap, arg->val.str.startmb, arg->val.str.lenmb);

  /* Determine the wide character.  */
  if (seq == NULL || seq->ucs4 == UNINITIALIZED_CHAR_VALUE)
    {
      wc = repertoire_find_value (repertoire, arg->val.str.startmb,
				  arg->val.str.lenmb);
      if (seq != NULL)
	seq->ucs4 = wc;
    }
  else
    wc = seq->ucs4;

  if (wc == ILLEGAL_CHAR_VALUE && seq == NULL)
    {
      /* It's no character, so look through the collation elements and
	 symbol list.  */
      void *result;

      if (find_entry (&collate->sym_table, arg->val.str.startmb,
		      arg->val.str.lenmb, &result) == 0)
	{
	  /* It's a collation symbol.  */
	  struct symbol_t *sym = (struct symbol_t *) result;
	  elem = sym->order;

	  if (elem == NULL)
	    elem = sym->order = new_element (collate, NULL, NULL, NULL, 0);
	}
      else if (find_entry (&collate->elem_table, arg->val.str.startmb,
			   arg->val.str.lenmb, (void **) &elem) != 0)
	{
	  /* It's also no collation element.  Therefore ignore it.  */
	  lr_ignore_rest (ldfile, 0);
	  return;
	}
    }
  else
    {
      /* Otherwise the symbols stands for a character.  */
      if (find_entry (&collate->seq_table, arg->val.str.startmb,
		      arg->val.str.lenmb, (void **) &elem) != 0)
	{
	  uint32_t wcs[2] = { wc, 0 };

	  /* We have to allocate an entry.  */
	  elem = new_element (collate, seq != NULL ? seq->bytes : NULL,
			      wcs, arg->val.str.startmb, arg->val.str.lenmb);

	  /* And add it to the table.  */
	  if (insert_entry (&collate->seq_table, arg->val.str.startmb,
			    arg->val.str.lenmb, elem) != 0)
	    /* This cannot happen.  */
	    assert (! "Internal error");
	}
    }

  /* Test whether this element is not already in the list.  */
  if (elem->next != NULL || (collate->cursor != NULL
			     && elem->next == collate->cursor))
    {
      lr_error (ldfile, _("order for `%.*s' already defined at %s:%Z"),
		arg->val.str.lenmb, arg->val.str.startmb,
		elem->file, elem->line);
      lr_ignore_rest (ldfile, 0);
      return;
    }

  insert_weights (ldfile, elem, charmap, repertoire, collate, tok_none);
}


static void
handle_ellipsis (struct linereader *ldfile, struct token *arg,
		 enum token_t ellipsis, struct charmap_t *charmap,
		 struct repertoire_t *repertoire,
		 struct locale_collate_t *collate)
{
  struct element_t *startp;
  struct element_t *endp;

  /* Unlink the entry added for the ellipsis.  */
  unlink_element (collate);
  startp = collate->cursor;

  /* Process and add the end-entry.  */
  if (arg != NULL)
    insert_value (ldfile, arg, charmap, repertoire, collate);

  /* Reset the cursor.  */
  collate->cursor = startp;

  /* Now we have to handle many different situations:
     - we have to distinguish between the three different ellipsis forms
     - the is the ellipsis at the beginning, in the middle, or at the end.
  */
  endp = collate->cursor->next;
  assert (arg == NULL || endp != NULL);

  /* Both, the start and the end symbol, must stand for characters.  */
  if ((startp == NULL || startp->name == NULL)
      || (endp == NULL || endp->name == NULL))
    {
      lr_error (ldfile, _("\
%s: the start end the end symbol of a range must stand for characters"),
		"LC_COLLATE");
      return;
    }

  if (ellipsis == tok_ellipsis3)
    {
      /* XXX */
    }
  else
    {
      /* For symbolic range we naturally must have a beginning and an
	 end specified by the user.  */
      if (startp == NULL)
	lr_error (ldfile, _("\
%s: symbolic range ellipsis must not directly follow `order_start'"),
		  "LC_COLLATE");
      else if (endp == NULL)
	lr_error (ldfile, _("\
%s: symbolic range ellipsis must not be direct followed by `order_end'"),
		  "LC_COLLATE");
      else
	{
	  /* Determine the range.  To do so we have to determine the
             common prefix of the both names and then the numeric
             values of both ends.  */
	  size_t lenfrom = strlen (startp->name);
	  size_t lento = strlen (endp->name);
	  char buf[lento + 1];
	  int preflen = 0;
	  long int from;
	  long int to;
	  char *cp;
	  int base = ellipsis == tok_ellipsis2 ? 16 : 10;

	  if (lenfrom != lento)
	    {
	    invalid_range:
	      lr_error (ldfile, _("\
`%s' and `%.*s' are no valid names for symbolic range"),
			startp->name, lento, endp->name);
	      return;
	    }

	  while (startp->name[preflen] == endp->name[preflen])
	    if (startp->name[preflen] == '\0')
	      /* Nothing to be done.  The start and end point are identical
		 and while inserting the end point we have already given
		 the user an error message.  */
	      return;
	    else
	      ++preflen;

	  errno = 0;
	  from = strtol (startp->name + preflen, &cp, base);
	  if ((from == UINT_MAX && errno == ERANGE) || *cp != '\0')
	    goto invalid_range;

	  errno = 0;
	  to = strtol (endp->name + preflen, &cp, base);
	  if ((to == UINT_MAX && errno == ERANGE) || *cp != '\0')
	    goto invalid_range;

	  /* Copy the prefix.  */
	  memcpy (buf, startp->name, preflen);

	  /* Loop over all values.  */
	  for (++from; from < to; ++from)
	    {
	      struct element_t *elem = NULL;
	      struct charseq *seq;
	      uint32_t wc;
	      int cnt;

	      /* Generate the the name.  */
	      sprintf (buf + preflen, base == 10 ? "%d" : "%x", from);

	      /* Look whether this name is already defined.  */
	      if (find_entry (&collate->seq_table, arg->val.str.startmb,
			      arg->val.str.lenmb, (void **) &elem) == 0)
		{
		  if (elem->next != NULL || (collate->cursor != NULL
					     && elem->next == collate->cursor))
		    {
		      lr_error (ldfile, _("\
%s: order for `%.*s' already defined at %s:%Z"),
				"LC_COLLATE", lenfrom, buf,
				elem->file, elem->line);
		      continue;
		    }

		  if (elem->name == NULL)
		    {
		      lr_error (ldfile, _("%s: `%s' must be a charater"),
				"LC_COLLATE", buf);
		      continue;
		    }
		}

	      if (elem == NULL || (elem->mbs == NULL && elem->wcs == NULL))
		{
		  /* Search for a character of this name.  */
		  seq = charmap_find_value (charmap, buf, lenfrom);
		  if (seq == NULL || seq->ucs4 == UNINITIALIZED_CHAR_VALUE)
		    {
		      wc = repertoire_find_value (repertoire, buf, lenfrom);

		      if (seq != NULL)
			seq->ucs4 = wc;
		    }
		  else
		    wc = seq->ucs4;

		  if (wc == ILLEGAL_CHAR_VALUE && seq == NULL)
		    /* We don't know anything about a character with this
		       name.  XXX Should we warn?  */
		    continue;

		  if (elem == NULL)
		    {
		      uint32_t wcs[2] = { wc, 0 };

		      /* We have to allocate an entry.  */
		      elem = new_element (collate,
					  seq != NULL ? seq->bytes : NULL,
					  wc == ILLEGAL_CHAR_VALUE
					  ? NULL : wcs,
					  buf, lenfrom);
		    }
		  else
		    {
		      /* Update the element.  */
		      if (seq != NULL)
			elem->mbs = obstack_copy0 (&collate->mempool,
						   seq->bytes, seq->nbytes);

		      if (wc != ILLEGAL_CHAR_VALUE)
			{
			  uint32_t zero = 0;

			  obstack_grow (&collate->mempool,
					&wc, sizeof (uint32_t));
			  obstack_grow (&collate->mempool,
					&zero, sizeof (uint32_t));
			  elem->wcs = obstack_finish (&collate->mempool);
			}
		    }

		  elem->file = ldfile->fname;
		  elem->line = ldfile->lineno;
		}

	      /* Enqueue the new element.  */
	      elem->last = collate->cursor;
	      elem->next = collate->cursor->next;
	      elem->last->next = elem;
	      if (elem->next != NULL)
		elem->next->last = elem;
	      collate->cursor = elem;

	      /* Now add the weights.  They come from the `ellipsis_weights'
		 member of `collate'.  */
	      elem->weights = (struct element_list_t *)
		obstack_alloc (&collate->mempool,
			       nrules * sizeof (struct element_list_t));
	      for (cnt = 0; cnt < nrules; ++cnt)
		if (collate->ellipsis_weight.weights[cnt].cnt == 1
		    && (collate->ellipsis_weight.weights[cnt].w[0]
			== ELEMENT_ELLIPSIS2))
		  {
		    elem->weights[cnt].w = (struct element_t **)
		      obstack_alloc (&collate->mempool,
				     sizeof (struct element_t *));
		    elem->weights[cnt].w[0] = elem;
		    elem->weights[cnt].cnt = 1;
		  }
		else
		  {
		    /* Simly use the weight from `ellipsis_weight'.  */
		    elem->weights[cnt].w =
		      collate->ellipsis_weight.weights[cnt].w;
		    elem->weights[cnt].cnt =
		      collate->ellipsis_weight.weights[cnt].cnt;
		  }
	    }
	}
    }
}


static void
collate_startup (struct linereader *ldfile, struct localedef_t *locale,
		 struct localedef_t *copy_locale, int ignore_content)
{
  if (!ignore_content)
    {
      struct locale_collate_t *collate;

      if (copy_locale == NULL)
	{
	  collate = locale->categories[LC_COLLATE].collate =
	    (struct locale_collate_t *)
	    xcalloc (1, sizeof (struct locale_collate_t));

	  /* Init the various data structures.  */
	  init_hash (&collate->elem_table, 100);
	  init_hash (&collate->sym_table, 100);
	  init_hash (&collate->seq_table, 500);
	  obstack_init (&collate->mempool);

	  collate->col_weight_max = -1;
	}
      else
	collate = locale->categories[LC_COLLATE].collate =
	  copy_locale->categories[LC_COLLATE].collate;
    }

  ldfile->translate_strings = 0;
  ldfile->return_widestr = 0;
}


void
collate_finish (struct localedef_t *locale, struct charmap_t *charmap)
{
}


void
collate_output (struct localedef_t *locale, struct charmap_t *charmap,
		const char *output_path)
{
}


void
collate_read (struct linereader *ldfile, struct localedef_t *result,
	      struct charmap_t *charmap, const char *repertoire_name,
	      int ignore_content)
{
  struct repertoire_t *repertoire = NULL;
  struct locale_collate_t *collate;
  struct token *now;
  struct token *arg = NULL;
  enum token_t nowtok;
  int state = 0;
  enum token_t was_ellipsis = tok_none;
  struct localedef_t *copy_locale = NULL;

  /* Get the repertoire we have to use.  */
  if (repertoire_name != NULL)
    repertoire = repertoire_read (repertoire_name);

  /* The rest of the line containing `LC_COLLATE' must be free.  */
  lr_ignore_rest (ldfile, 1);

  do
    {
      now = lr_token (ldfile, charmap, NULL);
      nowtok = now->tok;
    }
  while (nowtok == tok_eol);

  if (nowtok == tok_copy)
    {
      state = 2;
      now = lr_token (ldfile, charmap, NULL);
      if (now->tok != tok_string)
	{
	  SYNTAX_ERROR (_("%s: syntax error"), "LC_COLLATE");

	skip_category:
	  do
	    now = lr_token (ldfile, charmap, NULL);
	  while (now->tok != tok_eof && now->tok != tok_end);

	  if (now->tok != tok_eof
	      || (now = lr_token (ldfile, charmap, NULL), now->tok == tok_eof))
	    lr_error (ldfile, _("%s: premature end of file"), "LC_COLLATE");
	  else if (now->tok != tok_lc_collate)
	    {
	      lr_error (ldfile, _("\
%1$s: definition does not end with `END %1$s'"), "LC_COLLATE");
	      lr_ignore_rest (ldfile, 0);
	    }
	  else
	    lr_ignore_rest (ldfile, 1);

	  return;
	}

      /* Get the locale definition.  */
      copy_locale = find_locale (LC_COLLATE, now->val.str.startmb,
				 repertoire_name, charmap);
      if ((copy_locale->avail & COLLATE_LOCALE) == 0)
	{
	  /* Not yet loaded.  So do it now.  */
	  if (locfile_read (copy_locale, charmap) != 0)
	    goto skip_category;
	}

      lr_ignore_rest (ldfile, 1);

      now = lr_token (ldfile, charmap, NULL);
      nowtok = now->tok;
    }

  /* Prepare the data structures.  */
  collate_startup (ldfile, result, copy_locale, ignore_content);
  collate = result->categories[LC_COLLATE].collate;

  while (1)
    {
      /* Of course we don't proceed beyond the end of file.  */
      if (nowtok == tok_eof)
	break;

      /* Ingore empty lines.  */
      if (nowtok == tok_eol)
	{
	  now = lr_token (ldfile, charmap, NULL);
	  nowtok = now->tok;
	  continue;
	}

      switch (nowtok)
	{
	case tok_coll_weight_max:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 0)
	    goto err_label;

	  arg = lr_token (ldfile, charmap, NULL);
	  if (arg->tok != tok_number)
	    goto err_label;
	  if (collate->col_weight_max != -1)
	    lr_error (ldfile, _("%s: duplicate definition of `%s'"),
		      "LC_COLLATE", "col_weight_max");
	  else
	    collate->col_weight_max = arg->val.num;
	  lr_ignore_rest (ldfile, 1);
	  break;

	case tok_section_symbol:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 0)
	    goto err_label;

	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok != tok_bsymbol)
	    goto err_label;
	  else if (!ignore_content)
	    {
	      /* Check whether this section is already known.  */
	      struct section_list *known = collate->sections;
	      while (known != NULL)
		if (strcmp (known->name, arg->val.str.startmb) == 0)
		  break;

	      if (known != NULL)
		{
		  lr_error (ldfile,
			    _("%s: duplicate declaration of section `%s'"),
			    "LC_COLLATE", arg->val.str.startmb);
		  free (arg->val.str.startmb);
		}
	      else
		collate->sections = make_seclist_elem (collate,
						       arg->val.str.startmb,
						       collate->sections);

	      lr_ignore_rest (ldfile, known == NULL);
	    }
	  else
	    {
	      free (arg->val.str.startmb);
	      lr_ignore_rest (ldfile, 0);
	    }
	  break;

	case tok_collating_element:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 0)
	    goto err_label;

	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok != tok_bsymbol)
	    goto err_label;
	  else
	    {
	      const char *symbol = arg->val.str.startmb;
	      size_t symbol_len = arg->val.str.lenmb;

	      /* Next the `from' keyword.  */
	      arg = lr_token (ldfile, charmap, repertoire);
	      if (arg->tok != tok_from)
		{
		  free ((char *) symbol);
		  goto err_label;
		}

	      ldfile->return_widestr = 1;

	      /* Finally the string with the replacement.  */
	      arg = lr_token (ldfile, charmap, repertoire);
	      ldfile->return_widestr = 0;
	      if (arg->tok != tok_string)
		goto err_label;

	      if (!ignore_content)
		{
		  if (symbol == NULL)
		    lr_error (ldfile, _("\
%s: unknown character in collating element name"),
			      "LC_COLLATE");
		  if (arg->val.str.startmb == NULL)
		    lr_error (ldfile, _("\
%s: unknown character in collating element definition"),
			      "LC_COLLATE");
		  if (arg->val.str.startwc == NULL)
		    lr_error (ldfile, _("\
%s: unknown wide character in collating element definition"),
			      "LC_COLLATE");
		  else if (arg->val.str.lenwc < 2)
		    lr_error (ldfile, _("\
%s: substitution string in collating element definition must have at least two characters"),
			      "LC_COLLATE");

		  if (symbol != NULL)
		    {
		      /* The name is already defined.  */
		      if (check_duplicate (ldfile, collate, charmap,
					   repertoire, symbol, symbol_len))
			goto col_elem_free;

		      if (insert_entry (&collate->elem_table,
					symbol, symbol_len,
					new_element (collate,
						     NULL, NULL, NULL, 0)) < 0)
			lr_error (ldfile, _("\
error while adding collating element"));
		    }
		  else
		    goto col_elem_free;
		}
	      else
		{
		col_elem_free:
		  if (symbol != NULL)
		    free ((char *) symbol);
		  if (arg->val.str.startmb != NULL)
		    free (arg->val.str.startmb);
		  if (arg->val.str.startwc != NULL)
		    free (arg->val.str.startwc);
		}
	      lr_ignore_rest (ldfile, 1);
	    }
	  break;

	case tok_collating_symbol:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 0)
	    goto err_label;

	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok != tok_bsymbol)
	    goto err_label;
	  else
	    {
	      const char *symbol = arg->val.str.startmb;
	      size_t symbol_len = arg->val.str.lenmb;

	      if (!ignore_content)
		{
		  if (symbol == NULL)
		    lr_error (ldfile, _("\
%s: unknown character in collating symbol name"),
			      "LC_COLLATE");
		  else
		    {
		      /* The name is already defined.  */
		      if (check_duplicate (ldfile, collate, charmap,
					   repertoire, symbol, symbol_len))
			goto col_sym_free;

		      if (insert_entry (&collate->sym_table,
					symbol, symbol_len,
					new_symbol (collate)) < 0)
			lr_error (ldfile, _("\
error while adding collating symbol"));
		    }
		}
	      else
		{
		col_sym_free:
		  if (symbol != NULL)
		    free ((char *) symbol);
		}
	      lr_ignore_rest (ldfile, 1);
	    }
	  break;

	case tok_symbol_equivalence:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 0)
	    goto err_label;

	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok != tok_bsymbol)
	    goto err_label;
	  else
	    {
	      const char *newname = arg->val.str.startmb;
	      size_t newname_len = arg->val.str.lenmb;
	      const char *symname;
	      size_t symname_len;
	      struct symbol_t *symval;

	      arg = lr_token (ldfile, charmap, repertoire);
	      if (arg->tok != tok_bsymbol)
		{
		  if (newname != NULL)
		    free ((char *) newname);
		  goto err_label;
		}

	      symname = arg->val.str.startmb;
	      symname_len = arg->val.str.lenmb;

	      if (!ignore_content)
		{
		  if (newname == NULL)
		    {
		      lr_error (ldfile, _("\
%s: unknown character in equivalent definition name"),
				"LC_COLLATE");
		      goto sym_equiv_free;
		    }
		  if (symname == NULL)
		    {
		      lr_error (ldfile, _("\
%s: unknown character in equivalent definition value"),
				"LC_COLLATE");
		      goto sym_equiv_free;
		    }
		  /* The name is already defined.  */
		  if (check_duplicate (ldfile, collate, charmap,
				       repertoire, symname, symname_len))
		    goto col_sym_free;

		  /* See whether the symbol name is already defined.  */
		  if (find_entry (&collate->sym_table, symname, symname_len,
				  (void **) &symval) != 0)
		    {
		      lr_error (ldfile, _("\
%s: unknown symbol `%s' in equivalent definition"),
				"LC_COLLATE", symname);
		      goto col_sym_free;
		    }

		  if (insert_entry (&collate->sym_table,
				    newname, newname_len, symval) < 0)
		    {
		      lr_error (ldfile, _("\
error while adding equivalent collating symbol"));
		      goto sym_equiv_free;
		    }

		  free ((char *) symname);
		}
	      else
		{
		sym_equiv_free:
		  if (newname != NULL)
		    free ((char *) newname);
		  if (symname != NULL)
		    free ((char *) symname);
		}
	      lr_ignore_rest (ldfile, 1);
	    }
	  break;

	case tok_order_start:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 0 && state != 1)
	    goto err_label;
	  state = 1;

	  /* The 14652 draft does not specify whether all `order_start' lines
	     must contain the same number of sort-rules, but 14651 does.  So
	     we require this here as well.  */
	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok == tok_bsymbol)
	    {
	      /* This better should be a section name.  */
	      struct section_list *sp = collate->sections;
	      while (sp != NULL
		     && strcmp (sp->name, arg->val.str.startmb) != 0)
		sp = sp->next;

	      if (sp == NULL)
		{
		  lr_error (ldfile, _("\
%s: unknown section name `%s'"),
			    "LC_COLLATE", arg->val.str.startmb);
		  /* We use the error section.  */
		  collate->current_section = &collate->error_section;
		}
	      else
		{
		  /* Remember this section.  */
		  collate->current_section = sp;

		  /* One should not be allowed to open the same
                     section twice.  */
		  if (sp->first != NULL)
		    lr_error (ldfile, _("\
%s: multiple order definitions for section `%s'"),
			      "LC_COLLATE", sp->name);

		  /* Next should come the end of the line or a semicolon.  */
		  arg = lr_token (ldfile, charmap, repertoire);
		  if (arg->tok == tok_eol)
		    {
		      uint32_t cnt;

		      /* This means we have exactly one rule: `forward'.  */
		      if (collate->nrules > 1)
			lr_error (ldfile, _("\
%s: invalid number of sorting rules"),
				  "LC_COLLATE");
		      else
			collate->nrules = 1;
		      sp->rules = obstack_alloc (&collate->mempool,
						 (sizeof (enum coll_sort_rule)
						  * collate->nrules));
		      for (cnt = 0; cnt < collate->nrules; ++cnt)
			sp->rules[cnt] = sort_forward;

		      /* Next line.  */
		      break;
		    }

		  /* Get the next token.  */
		  arg = lr_token (ldfile, charmap, repertoire);
		}
	    }
	  else
	    {
	      /* There is no section symbol.  Therefore we use the unnamed
		 section.  */
	      collate->current_section = &collate->unnamed_section;

	      if (collate->unnamed_section.first != NULL)
		lr_error (ldfile, _("\
%s: multiple order definitions for unnamed section"),
			  "LC_COLLATE");
	    }

	  /* Now read the direction names.  */
	  read_directions (ldfile, arg, charmap, repertoire, collate);

	  /* From now be need the strings untranslated.  */
	  ldfile->translate_strings = 0;
	  break;

	case tok_order_end:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 1)
	    goto err_label;

	  /* Handle ellipsis at end of list.  */
	  if (was_ellipsis)
	    /* XXX */
	    abort ();

	  state = 2;
	  lr_ignore_rest (ldfile, 1);
	  break;

	case tok_reorder_after:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state == 1)
	    {
	      lr_error (ldfile, _("%s: missing `order_end' keyword"),
			"LC_COLLATE");
	      state = 2;

	      /* Handle ellipsis at end of list.  */
	      if (was_ellipsis)
		/* XXX */
		abort ();
	    }
	  else if (state != 2 && state != 3)
	    goto err_label;
	  state = 3;

	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok == tok_bsymbol)
	    {
	      /* Find this symbol in the sequence table.  */
	      struct element_t *insp;
	      int no_error = 1;

	      if (find_entry (&collate->seq_table, arg->val.str.startmb,
			      arg->val.str.lenmb, (void **) &insp) == 0)
		/* Yes, the symbol exists.  Simply point the cursor
		   to it.  */
		  collate->cursor = insp;
	      else
		{
		  /* This is bad.  The symbol after which we have to
                     insert does not exist.  */
		  lr_error (ldfile, _("\
%s: cannot reorder after %.*s: symbol not known"),
			    "LC_COLLATE", arg->val.str.lenmb,
			    arg->val.str.startmb);
		  collate->cursor = NULL;
		  no_error = 0;
		}

	      lr_ignore_rest (ldfile, no_error);
	    }
	  else
	    /* This must not happen.  */
	    goto err_label;
	  break;

	case tok_reorder_end:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    break;

	  if (state != 3)
	    goto err_label;
	  state = 4;
	  lr_ignore_rest (ldfile, 1);
	  break;

	case tok_reorder_sections_after:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state == 1)
	    {
	      lr_error (ldfile, _("%s: missing `order_end' keyword"),
			"LC_COLLATE");
	      state = 2;

	      /* Handle ellipsis at end of list.  */
	      if (was_ellipsis)
		/* XXX */
		abort ();
	    }
	  else if (state == 3)
	    {
	      error (0, 0, _("%s: missing `reorder-end' keyword"),
		     "LC_COLLATE");
	      state = 4;
	    }
	  else if (state != 2 && state != 4)
	    goto err_label;
	  state = 5;

	  /* Get the name of the sections we are adding after.  */
	  arg = lr_token (ldfile, charmap, repertoire);
	  if (arg->tok == tok_bsymbol)
	    {
	      /* Now find a section with this name.  */
	      struct section_list *runp = collate->sections;

	      while (runp != NULL)
		{
		  if (runp->name != NULL
		      && strlen (runp->name) == arg->val.str.lenmb
		      && memcmp (runp->name, arg->val.str.startmb,
				 arg->val.str.lenmb) == 0)
		    break;

		  runp = runp->next;
		}

	      if (runp != NULL)
		collate->current_section = runp;
	      else
		{
		  /* This is bad.  The section after which we have to
                     reorder does not exist.  Therefore we cannot
                     process the whole rest of this reorder
                     specification.  */
		  lr_error (ldfile, _("%s: section `%.*s' not known"),
			    "LC_COLLATE", arg->val.str.lenmb,
			    arg->val.str.startmb);

		  do
		    {
		      lr_ignore_rest (ldfile, 0);

		      now = lr_token (ldfile, charmap, NULL);
		    }
		  while (now->tok == tok_reorder_sections_after
			 || now->tok == tok_reorder_sections_end
			 || now->tok == tok_end);

		  /* Process the token we just saw.  */
		  nowtok = now->tok;
		  continue;
		}
	    }
	  else
	    /* This must not happen.  */
	    goto err_label;
	  break;

	case tok_reorder_sections_end:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    break;

	  if (state != 5)
	    goto err_label;
	  state = 6;
	  lr_ignore_rest (ldfile, 1);
	  break;

	case tok_bsymbol:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 1 && state != 3)
	    goto err_label;

	  if (state == 3)
	    {
	      /* It is possible that we already have this collation sequence.
		 In this case we move the entry.  */
	      struct element_t *seqp;

	      /* If the symbol after which we have to insert was not found
		 ignore all entries.  */
	      if (collate->cursor == NULL)
		{
		  lr_ignore_rest (ldfile, 0);
		  break;
		}

	      if (find_entry (&collate->seq_table, arg->val.str.startmb,
			      arg->val.str.lenmb, (void **) &seqp) == 0)
		{
		  /* Remove the entry from the old position.  */
		  if (seqp->last == NULL)
		    collate->start = seqp->next;
		  else
		    seqp->last->next = seqp->next;
		  if (seqp->next != NULL)
		    seqp->next->last = seqp->last;

		  /* We also have to check whether this entry is the
                     first or last of a section.  */
		  if (seqp->section->first == seqp)
		    {
		      if (seqp->section->first == seqp->section->last)
			/* This setion has no content anymore.  */
			seqp->section->first = seqp->section->last = NULL;
		      else
			seqp->section->first = seqp->next;
		    }
		  else if (seqp->section->last == seqp)
		    seqp->section->last = seqp->last;

		  /* Now insert it in the new place.  */
		  seqp->next = collate->cursor->next;
		  seqp->last = collate->cursor;
		  collate->cursor->next = seqp;
		  if (seqp->next != NULL)
		    seqp->next->last = seqp;

		  seqp->section = collate->cursor->section;
		  if (seqp->section->last == collate->cursor)
		    seqp->section->last = seqp;

		  break;
		}

	      /* Otherwise we just add a new entry.  */
	    }
	  else if (state == 5)
	    {
	      /* We are reordering sections.  Find the named section.  */
	      struct section_list *runp = collate->sections;
	      struct section_list *prevp = NULL;

	      while (runp != NULL)
		{
		  if (runp->name != NULL
		      && strlen (runp->name) == arg->val.str.lenmb
		      && memcmp (runp->name, arg->val.str.startmb,
				 arg->val.str.lenmb) == 0)
		    break;

		  prevp = runp;
		  runp = runp->next;
		}

	      if (runp == NULL)
		{
		  lr_error (ldfile, _("%s: section `%.*s' not known"),
			    "LC_COLLATE", arg->val.str.lenmb,
			    arg->val.str.startmb);
		  lr_ignore_rest (ldfile, 0);
		}
	      else
		{
		  if (runp != collate->current_section)
		    {
		      /* Remove the named section from the old place and
			 insert it in the new one.  */
		      prevp->next = runp->next;

		      runp->next = collate->current_section->next;
		      collate->current_section->next = runp;
		      collate->current_section = runp;
		    }

		  /* Process the rest of the line which might change
                     the collation rules.  */
		  arg = lr_token (ldfile, charmap, repertoire);
		  if (arg->tok != tok_eof && arg->tok != tok_eol)
		    read_directions (ldfile, arg, charmap, repertoire,
				     collate);
		}
	      break;
	    }
	  else if (was_ellipsis != tok_none)
	    {
	      /* Using the information in the `ellipsis_weight'
                 element and this and the last value we have to handle
                 the ellipsis now.  */
	      assert (state == 1);

	      handle_ellipsis (ldfile, arg, was_ellipsis, charmap, repertoire,
			       collate);

	      /* Remember that we processed the ellipsis.  */
	      was_ellipsis = tok_none;

	      /* And don't add the value a second time.  */
	      break;
	    }

	  /* Now insert in the new place.  */
	  insert_value (ldfile, arg, charmap, repertoire, collate);
	  break;

	case tok_undefined:
	  /* Ignore the rest of the line if we don't need the input of
	     this line.  */
	  if (ignore_content)
	    {
	      lr_ignore_rest (ldfile, 0);
	      break;
	    }

	  if (state != 1)
	    goto err_label;

	  if (was_ellipsis != tok_none)
	    {
	      lr_error (ldfile,
			_("%s: cannot have `%s' as end of ellipsis range"),
			"LC_COLLATE", "UNDEFINED");

	      unlink_element (collate);
	      was_ellipsis = tok_none;
	    }

	  /* See whether UNDEFINED already appeared somewhere.  */
	  if (collate->undefined.next != NULL
	      || (collate->cursor != NULL
		  && collate->undefined.next == collate->cursor))
	    {
	      lr_error (ldfile,
			_("%s: order for `%.*s' already defined at %s:%Z"),
			"LC_COLLATE", 9, "UNDEFINED", collate->undefined.file,
			collate->undefined.line);
	      lr_ignore_rest (ldfile, 0);
	    }
	  else
	    /* Parse the weights.  */
	     insert_weights (ldfile, &collate->undefined, charmap,
			     repertoire, collate, tok_none);
	  break;

	case tok_ellipsis2:
	case tok_ellipsis3:
	case tok_ellipsis4:
	  /* This is the symbolic (decimal or hexadecimal) or absolute
             ellipsis.  */
	  if (was_ellipsis != tok_none)
	    goto err_label;

	  if (state != 1 && state != 3)
	    goto err_label;

	  was_ellipsis = nowtok;

	  insert_weights (ldfile, &collate->ellipsis_weight, charmap,
			  repertoire, collate, nowtok);
	  break;

	case tok_end:
	  /* Next we assume `LC_COLLATE'.  */
	  if (!ignore_content)
	    {
	      if (state == 0)
		/* We must either see a copy statement or have
		   ordering values.  */
		lr_error (ldfile,
			  _("%s: empty category description not allowed"),
			  "LC_COLLATE");
	      else if (state == 1)
		{
		  lr_error (ldfile, _("%s: missing `order_end' keyword"),
			    "LC_COLLATE");

		  /* Handle ellipsis at end of list.  */
		  if (was_ellipsis)
		    /* XXX */
		    abort ();
		}
	      else if (state == 3)
		error (0, 0, _("%s: missing `reorder-end' keyword"),
		       "LC_COLLATE");
	      else if (state == 5)
		error (0, 0, _("%s: missing `reorder-sections-end' keyword"),
		       "LC_COLLATE");
	    }
	  arg = lr_token (ldfile, charmap, NULL);
	  if (arg->tok == tok_eof)
	    break;
	  if (arg->tok == tok_eol)
	    lr_error (ldfile, _("%s: incomplete `END' line"), "LC_COLLATE");
	  else if (arg->tok != tok_lc_collate)
	    lr_error (ldfile, _("\
%1$s: definition does not end with `END %1$s'"), "LC_COLLATE");
	  lr_ignore_rest (ldfile, arg->tok == tok_lc_collate);
	  return;

	default:
	err_label:
	  SYNTAX_ERROR (_("%s: syntax error"), "LC_COLLATE");
	}

      /* Prepare for the next round.  */
      now = lr_token (ldfile, charmap, NULL);
      nowtok = now->tok;
    }

  /* When we come here we reached the end of the file.  */
  lr_error (ldfile, _("%s: premature end of file"), "LC_COLLATE");
}